INTRODUCTIONThe Adenosine Axis in Cancer

• Standard chemotherapy regimens may contribute to immunosuppression by elevating intratumoral levels of adenosine triphosphate (ATP) in the tumor microenvironment (TME) where the enzymes CD39 and CD73 successively convert ATP to adenosine1,2 (Figure 1)

• By binding adenosine receptors 2a and 2b (A2aR and A2bR) expressed on immune cells, adenosine promotes immunosuppression by inhibiting critical components of the antitumor immune response, ultimately enabling tumors to evade destruction2

• Additionally, A2aR signaling impairs the activation, proliferation, and cytotoxic activity of effector T cells3

• Initial research focused on A2aR as the most relevant adenosine receptor in cancer physiology; however, A2bR signaling through MAP kinase pathway activation mediates unique functions, such as cancer cell intrinsic survival and dendritic cell activation and function4

• Thus, adenosine pathway blockade may be necessary to overcome adenosine-dependent immunosuppression leading to enhanced therapeutic efficacy of some chemotherapeutic agents2

Figure 1. Critical Role of Adenosine Pathway in the Immunosuppressive TME

AMP, adenosine monophosphate; ATP, adenosine triphosphate; A2aR/A2bR, adenosine receptors 2a/2b; DC, dendritic cell; IL, interleukin; MDSC, myeloid-derived suppressor cell; NK, natural killer; PD-1, programmed cell death protein-1; TAM, tumor-associated macrophage; TME, tumor microenvironment; TNAP, tissue nonspecific alkaline phosphatase.

• AB680 is a potent, selective small-molecule inhibitor of soluble and membrane-bound CD73 developed with the aim of eliminating adenosine-mediated immunosuppression in the TME; it is the first clinical-stage small-molecule CD73 inhibitor

• Preliminary data indicated that systemic exposure (Cmax and AUC) of 25-75 mg AB680 administered intravenously (IV) once every 2 weeks (Q2W) increased in a dose-proportional manner; AB680 pharmacokinetics (PK) profiles were similar in healthy volunteers and patients with cancer5

• AB680 administration to healthy volunteers at doses as low as 16 mg results in ~100% inhibition of circulating soluble CD73 for an extended period of time (>7 days), thus, the starting AB680 dose level in ARC-8 (25 mg Q2W) is known to be pharmacologically active and to provide close to maximal inhibition of CD73 in circulation6

ARC-8 Study Rationale• Standard treatment for metastatic pancreatic ductal adenocarcinoma (mPDAC) includes combination chemotherapy

regimens such as nab-paclitaxel and gemcitabine (NP/Gem); however, the overall and complete response rates for patients with mPDAC are 23-29% and 1%, respectively, with a 9% 5-year survival rate7,8

• Additionally, combinations of NP/Gem + programmed cell death protein-1 (PD-1) inhibitors have demonstrated an 18-27% overall response rate (ORR)9,10, further highlighting an urgent unmet need for new targeted treatment options

• Human pancreatic tumors express high levels of CD73, which are associated with KRAS mutation5 (Figure 2)

• Over 90% of invasive PDACs have mutated KRAS; compared with patients who have PDAC and wild-type KRAS, those with mutated KRAS have worse clinical outcomes6,11

• Higher CD73 expression is strongly associated with worse progression-free survival in patients with PDAC, regardless of KRAS status12

• In a murine B16F10 melanoma model, tumor growth was suppressed with AB680 monotherapy, an effect that was further enhanced with the addition of anti–PD-1 antibodies5

– AB680 + anti–PD-1 treatment increased the number of intratumoral effector CD4+ and CD8+ T cells and decreased the numbers of Tregs and myeloid-derived suppressor cells5

– Further, AMP-mediated suppression of human CD4+ and CD8+ T cell activation was reversed with AB680 treatment5

• Based on these observations, novel combinations of AB680 and PD-1 pathway inhibition may potentially offer a complementary approach to increase antitumor immune activation and improved anticancer therapy

Figure 2. KRAS Mutation Correlates With Elevated CD73 Expression

(A) Linear model estimates adjusted for tumor type of alterations in cancer driver genes that predict CD73 expression (B) Pancreatic adenocarcinoma samples from the TCGA demonstrate that tumors harboring KRAS mutations have significantly elevated CD73 expression levels (****p<0.001). FDR, false discovery rate; MUT, mutated, WT, wild type.

METHODSStudy Design

• ARC-8 (NCT04104672) is an ongoing, Phase 1/1b, open-label, dose-escalation, and dose-expansion study to evaluate the safety, tolerability, PK and clinical activity of AB680 in combination with NP/Gem and zimberelimab (Zim; AB122; anti–PD-1) as a first-line treatment for patients with mPDAC (Figure 3)

• The dose-escalation stage consists of a standard 3+3 design in which AB680 (25, 50, 75, or 100 mg) is administered IV Q2W with standard doses of NP/Gem + Zim (240 mg) to determine the recommended dose for expansion (RDE)

• In the dose-expansion, AB680 will be administered at the RDE with NP/Gem + Zim

Figure 3. ARC-8 Study Design

1L, first-line; IV, intravenously; Gem, gemcitabine; NP, nab-paclitaxel; PDAC, pancreatic ductal adenocarcinoma; Q2W, every 2 weeks; R, randomization; RDE, recommended dose for expansion; Zim, zimberelimab.

ARC-8 Design Features• The primary objective of the study is to assess safety and tolerability of AB680 combination therapy in patients with

mPDAC with secondary objectives that include AB680 PK and clinical activity

• Eligible patients are adults who have histologically- or cytologically-confirmed mPDAC and have not previously received treatment for metastatic disease; ≥1 measurable lesion per RECIST v1.1; and an ECOG performance status of 0-1

– Prior (neo)adjuvant treatment for PDAC (chemotherapy [including NP or Gem] and/or radiotherapy) is allowed if it was completed ≥6 months prior to study enrollment

– Patients initially diagnosed with locally advanced PDAC who have undergone chemotherapy then resection and had no evidence of disease are eligible for the study if relapse of metastatic disease has occurred and the last dose of chemotherapy was received ≥6 months prior to study enrollment

• Baseline archival tumor samples (≤12 months old) or on-treatment biopsies (if medically feasible) are collected from all patients

• Study treatment may continue until disease progression, unacceptable toxicity, consent withdrawal, or by the investigator’s decision

Statistical Analysis• Safety analyses included all patients who received ≥1 dose of any study drug; summary statistics are shown for treatment-

emergent adverse events (TEAEs), serious adverse events (TESAEs), event severity, and relationship to study drugs

• Efficacy analyses included all patients who had a baseline and ≥1 post-baseline assessment or discontinued study treatment due to progressive disease or death; patients were considered efficacy-evaluable if they had ≥1 disease assessment

RESULTSPatient Baseline Characteristics

• As of November 11, 2020, 19 patients have received AB680 + NP/Gem + Zim: 25 mg AB680 (n=4), 50 mg AB680 (n=6), 75 mg AB680 (n=3), and 100 mg AB680 (n=6; Table 1)

– Based on the totality of the data including PK, PK/PD correlation, and available safety data, 100 mg AB680 has been tentatively selected as the RDE

• The majority of study patients were men; the mean age of all study patients was 64 years

Table 1. Patient Demographics and Characteristics

Parameter

25 mg AB680 + NP/Gem + Zim

(n=4)

50 mg AB680 + NP/Gem + Zim

(n=6)

75 mg AB680 + NP/Gem + Zim

(n=3)

100 mg AB680 + NP/Gem + Zim

(n=6)All Patients

(N=19)

Age, mean (SD), years 67.3 (2.2) 61.8 (12.7) 55.3 (2.1) 68.5 (8.7) 64.1 (9.5)

Male, n (%) 2 (50) 4 (67) 2 (67) 3 (50) 11 (58)

Race, n (%) White Asian Not reported

4 (100)00

4 (67)1 (17)1 (17)

2 (67)1 (33)

0

4 (67)1 (17)1 (17)

14 (74)3 (16)2 (11)

ECOG PS, n (%) 0 1

4 (100)0

4 (67)2 (33)

3 (100)0

3 (50)3 (50)

14 (74)5 (26)

ECOG PS, Eastern Cooperative Oncology Group performance status; NP/Gem, nab-paclitaxel/gemcitabine; SD, standard deviation; Zim, zimberelimab.

Safety Analyses• As of November 11, 2020, 1 dose-limiting toxicity (Grade 2 autoimmune hepatitis) occurred in the 50 mg AB680 cohort;

the event resolved completely with steroid treatment and the patient resumed study treatment

• Of all safety-evaluable patients, 18/19 (95%) experienced ≥1 TEAE; the most common TEAEs were fatigue (n=13; 68%), anemia (n=10; 53%), alopecia (n=8; 42%), diarrhea (n=8; 42%), and neutrophil count decreased (n=8; 42%; Table 2)

• Study drug-related TEAEs were reported by 8/19 (42%) patients; 4 patients had Grade 1 or 2 study-drug related TEAEs and 4 patients experienced Grade 3 study drug-related TEAEs (n=2 in 50 mg AB680 cohort; n=1 each in 75 mg and 100 mg AB680 cohorts)

– Grade 3 study drug-related TEAEs: anemia (n=2), neutrophil count decreased (n=2), lymphocyte count decreased (n=1), platelet count decreased (n=1), and C. difficile colitis (n=1)

• TESAEs were reported by 7/19 (37%) patients; for 2 patients, these TESAEs experienced were study drug-related (Grade 2 blood bilirubin increased [n=1 in 50 mg AB680 cohort] and Grade 3 C. difficile colitis [n=1 in 100 mg AB680 cohort])

• There were no patients who discontinued study treatment due to TEAEs

• In limited patient numbers across each dose-escalation cohort, no significant additive toxicity has been observed over that expected with NP/Gem alone

Table 2. Treatment Emergent Adverse Events

Parameter, n (%)

25 mg AB680 + NP/Gem + Zim

(n=4)

50 mg AB680 + NP/Gem + Zim

(n=6)

75 mg AB680 + NP/Gem + Zim

(n=3)

100 mg AB680 + NP/Gem + Zim

(n=6)

All Patients(N=19)

Any TEAE 4 (100) 6 (100) 3 (100) 5 (83) 18 (95)

Grade ≥3 3 (75) 5 (83) 1 (33) 3 (50) 12 (63)

Any TESAE 1 (25) 3 (50) 0 3 (50) 7 (37)

Grade ≥3 1 (25) 2 (33) 0 3 (50) 6 (32)

Study drug-related TEAEs 1 (25) 3 (50) 2 (67) 2 (33) 8 (42)

Grade ≥3 0 2 (33) 1 (33) 1 (17) 4 (21)

Study drug-related TESAEs 0 1 (17) 0 1 (17) 2 (11)

Grade ≥3 0 0 0 1 (17) 1 (5)

TEAEs in ≥4 patients

Fatigue 4 (100) 6 (100) 1 (33) 2 (33) 13 (68)

Anemia 1 (25) 5 (83) 2 (67) 2 (33) 10 (53)

Alopecia 2 (50) 2 (33) 2 (67) 2 (33) 8 (42)

Diarrhea 2 (50) 4 (67) 1 (33) 1 (17) 8 (42)

Neutrophil count decreased 2 (50) 4 (67) 1 (33) 1 (17) 8 (42)

Nausea 1 (25) 3 (50) 2 (67) 1 (17) 7 (37)

Platelet count decreased 2 (50) 2 (33) 1 (33) 2 (33) 7 (37)

Pyrexia 3 (75) 3 (50) 1 (33) 0 7 (37)

AST increased 1 (25) 2 (33) 2 (67) 1 (17) 6 (32)

Vomiting 2 (50) 3 (50) 1 (33) 0 6 (32)

ALT increased 0 2 (33) 2 (67) 1 (17) 5 (26)

Blood ALP increased 0 1 (17) 2 (67) 2 (33) 5 (26)

Constipation 1 (25) 3 (50) 1 (33) 0 5 (26)

Lymphocyte count decreased 0 1 (17) 1 (33) 3 (50) 5 (26)

Chills 1 (25) 1 (17) 1 (33) 1 (17) 4 (21)

Decreased appetite 0 2 (33) 0 2 (33) 4 (21)

Hypoalbuminemia 0 1 (17) 1 (33) 2 (33) 4 (21)

ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; NP/Gem, nab-paclitaxel/gemcitabine; TEAE, treatment-emergent adverse event; TESAE, treatment-emergent serious adverse event; Zim, zimberelimab.

Clinical Activity• As of December 9, 2020, there were 17 efficacy-evaluable patients; partial responses (PR) were observed in 7 patients (41%),

including 3 confirmed responses (Figure 4); of the 4 unconfirmed responders, 3 responded at the first tumor assessment, and all 4 remain on study treatment

• As of December 9, 2020, 16 patients remained on active treatment; with limited follow-up, 6/6 patients receiving 100 mg AB680 (cohort 4) continue on therapy

Figure 4. Time on Treatment and RECIST v1.1 Response

IV, intravenously; NP/Gem, nab-paclitaxel/gemcitabine; PD, progressive disease; PR, partial response; Q2W, every 2 weeks; Zim, zimberelimab.

• Changes in target lesions over time are shown in Figures 5 and 6

Figure 5. Waterfall Plot of Best Percent Change from Baseline in Sum of Target Lesions

01−003 11−001 03−001 01−006 11−004 01−004 07−003 07−002 03−003 03−002 07−00406−001 08−00102−001 01−007 07−005 07−001

Sum

of T

arge

t Les

ion

Best

Per

cent

Cha

nge

From

Bas

elin

e

−100

−50

0

50

100AB680 25 mg IV Q2W + NP/Gem + ZimAB680 50 mg IV Q2W + NP/Gem + ZimAB680 75 mg IV Q2W + NP/Gem + ZimAB680 100 mg IV Q2W + NP/Gem + Zim

IV, intravenously; NP/Gem, nab-paclitaxel/gemcitabine; Q2W, every 2 weeks; Zim, zimberelimab.

Figure 6. Spider Plot of Percent Change from Baseline in Sum of Target Lesions

0 2 4 6 8

−100

−50

0

50

100

Study Month

Sum

of T

arge

t Les

ion

Perc

ent C

hang

e Fr

om B

asel

ine

AB680 25 mg IV Q2W + NP/Gem + Zim AB680 50 mg IV Q2W + NP/Gem+ ZimAB680 75 mg IV Q2W + NP/Gem + ZimAB680 100 mg IV Q2W + NP/Gem + ZimOff−treatment

IV, intravenously; NP/Gem, nab-paclitaxel/gemcitabine; Q2W, every 2 weeks; Zim, zimberelimab.

• Figure 7 shows radiologic images and CA19-9 levels over time for a 60 year old male patient who received his first AB680 dose (50 mg IV Q2W) on May 11, 2020. After Cycle 1, immunotherapy was interrupted due to Grade 2 autoimmune hepatitis, which resolved after pulse steroid therapy. Treatment with all 4 drugs resumed at the start of Cycle 3. Tumor shrinkage was observed across the patient’s 3 target lesions and CA19-9 levels declined precipitously after Cycle 2.

Figure 7. Patient 02-001: Radiological Disease Evaluation and CA19-9 Levels

C, cycle.

CONCLUSIONS• Preliminary results from ARC-8 indicate that AB680, the first clinical-stage small-molecule CD73 inhibitor, in combination

with standard-of-care chemotherapy + Zim has a manageable safety profile consistent with that expected for each agent alone and demonstrates early signals of clinical activity

• Combination treatment in 17 efficacy-evaluable patients resulted in a 41% overall response rate (7/17) which compares favorably with current standard-of-care chemotherapy (23-29% ORR for NP/Gem)

• Enrollment into the 100 mg AB680 (Cohort 4) is ongoing to inform selection of the RDE; initiation of the dose-expansion stage is expected in December 2020, followed by a 2:1 randomization evaluating AB680 + NP/Gem + Zim vs AB680 + NP/Gem pending futility analysis

ACKNOWLEDGMENTS AND DISCLOSURESThe authors gratefully acknowledge the patients, their families, and their caregivers for their participation in this clinical trial. They would also like to thank the ARC-8 Principal Investigators and study staff for carrying out the study. GA Manji has been a consultant/advisory board member for Ardelyx, BioLineRx, and Roche and has received research funding from BioLineRx, Merck, Regeneron, and Roche/Genentech. ZA Wainberg has been a consultant/advisory board member for Array BioPharma, AstraZeneca/MedImmune, Bayer, Bristol Myers Squibb, Eli Lilly, Five Prime Therapeutics, Ipsen, Merck & Co., Merck KGaA, Macrogenics, and Novartis; and has received research funding from Five Prime Therapeutics, Merck & Co, Novartis, Pfizer, and Plexxikon. K Krishnan, N Giafis, A Udyavar, CS Quah, S Liu, W Berry, D DiRenzo, K Gerrick, and L Jin are employees of Arcus Biosciences, Inc. and may own company stock. J Bendell has been a consultant for Amgen, Apexigen, Arch Oncology, ARMO BioSciences, Array BioPharma, AstraZeneca, Bayer, Beigene, Bicycle Therapeutics, Boehringer Ingelheim, Bristol Myers Squibb, Celgene, Cerulean Pharma, Continuum Clinical, Cyteir, Daiichi Sankyo, EMD Serono, Evelo Therapeutics, Five Prime Therapeutics, FORMA Therapeutics, Genentech/Roche, Gilead Sciences, GlaxoSmithKline, Incyte, Innate Pharma, Ipsen, Janssen, Kyn Therapeutics, Leap Therapeutics, Lilly, Macrogenics, MedImmune, Merck, Merrimack, Moderna Therapeutics, Molecular Partners, Novartis, Oncogenex, OncoMed, Phoenix Biotech, Prelude Therapeutics, Relay Therapeutics, Sanofi, Seattle Genetics, Taiho Pharmaceutical, Tanabe Research, TD2, TG Therapeutics, Tizona Therapeutics, Inc., Tolero Pharmaceuticals, and Torque; and has received research funding from Abbott/AbbVie, Acerta Pharma, ADC Therapeutics, Agios, Amgen, Apexigen, Arch Oncology, Arcus Biosciences, ARMO BioSciences, Array BioPharma, Arrys Therapeutics, AstraZeneca/MedImmune, AtlasMedx, Bayer, BeiGene, Bellicum Pharmaceuticals, Bicycle Therapeutics, Blueprint Medicines, Boehringer Ingelheim, Boston Biomedical, Bristol Myers Squibb, Calithera Biosciences, Celgene, Celldex, Cyteir, CytomX Therapeutics, Daiichi Sankyo, eFFECTOR Therapeutics, Eisai, EMD Serono, Evelo Therapeutics, Ipsen, Five Prime Therapeutics, FORMA Therapeutics, Forty Seven, Genentech/Roche, Gilead Sciences, GlaxoSmithKline, Gossamer Bio, Gritstone Oncology, Harpoon Therapeutics, Hutchison MediPharma, IGM, Incyte, Innate Pharma, ImClone Systems, Jacobio, Janssen, Kolltan Pharmaceuticals, Leap Therapeutics, Lilly, MabSpace Biosciences, Macrogenics, Marshall Edwards, Merck, Merrimack, Mersana, Merus, Millenium Pharmaceuticals, Molecular Partners, Nektar, NGM Biopharmaceuticals, Novartis, Novocure, Numab, Oncogenex, OncoMed, Onyx, Pfizer, Phoenix Biotech, Pieris Pharmaceuticals, Prelude Therapeutics, Relay Therapeutics, Repare Therapeutics, Revolution Medicines, Rgenix, Sanofi, Scholar Rock, Seattle Genetics, Shattuck Labs, Sierra Oncology, Sorrento Therapeutics, Stem CentRx, SynDevRx, Synthorx, Taiho Pharmaceutical, Takeda, Tarveda Therapeutics, Tempest Therapeutics, TG Therapeutics, Tizona Therapeutics, Inc., Torque, TRACON Pharma, Treadwell Therapeutics, Tyrogenex, Unum Therapeutic, Vyriad, and Zymeworks. Medical writing support was provided by Amanda Martin, PhD, of Medical Expressions (Chicago, IL) and was funded by Arcus Biosciences, Inc.

REFERENCES1. Martins I, et al. Cell Cycle. 2009;8(22):3723-3728.2. Vijayan D, et al. Nat Rev Cancer. 2017;17(12):709-724.3. Cekic C, et al. Nat Rev Immunol. 2016;16(3):177-192.4. Gao ZG and Jacobson KA. Int J Mol Sci. 2019;20(20):5139.5. Bendell J, et al. ASCO GI 2020. Abstract TPS788. 6. Ashok D, et al. SITC 2019. Abstract P379.

7. Von Hoff DD, et al. N Eng J Med. 2013;369(18):1691-1703.8. Orth M, et al. Radiat Oncol. 2019;14(1):141.9. Singh RR and O’Reilly EM. Drugs. 2020;80(7):647-669.

10. Wainberg ZA, et al. Clin Cancer Res. 2020;26(18):4814-4822.11. Qian ZR, et al. JAMA Oncol. 2018;4(3):e173420.12. Udyavar A, et al., unpublished data.

A2aR

A2aR

A2aRA2bR

Cytotoxicity

NK cells

T cells

DC, TAM, MDSC

Effector function Cytotoxicity PD-1

PD-1

IL-12 IL-10 T-cell stimulation

ATP

ATP released from dying tumor cells

Adenosine binding to A2aR/A2bR inhibits antitumor immune response

ATPATP AMP

Adenosine

CD39 CD73TNAP

Conversion of ATP to adenosine by CD39 and CD73

01−00302−00307−00506−00101−00407−00408−00303−00311−00403−00207−00308−00103−00101−00707−00202−00111−00101−00607−001

0 2 4 6

Duration of Treatment, Months

8 10

PD

PD

Clinical Progression

AB680 25 mg IV Q2W + NP/Gem + Zim AB680 50 mg IV Q2W + NP/Gem + Zim AB680 75 mg IV Q2W + NP/Gem + Zim AB680 100 mg IV Q2W + NP/Gem + Zim Death

PRPD

50mm 26mm

VisitPre-C1 Pre-C2 Pre-C4 Pre-C5 Pre-C6 Pre-C7

Res

ult

20000

15000

10000

5000

Apr 22 2020 Jun 08 2020 Aug 03 2020 Aug 31 2020 Sep 28 2020 Oct 12 2020 Oct 26 2020

0

Pre-Cycle 7Screening

ARC-8: Phase 1/1b Study to Evaluate Safety and Tolerability of AB680 + Chemotherapy + Zimberelimab (AB122) in Patients with Treatment-Naive Metastatic Pancreatic Adenocarcinoma

GA Manji1, ZA Wainberg2, K Krishnan3, N Giafis3, A Udyavar3, CS Quah3, S Liu3, W Berry3, D DiRenzo3, K Gerrick3, L Jin3, JC Bendell41Columbia University Herbert Irving Comprehensive Cancer Center, New York, NY; 2University of California Los Angeles David Geffen School of Medicine, Los Angeles, CA; 3Arcus Biosciences, Inc., Hayward, CA; 4Sarah Cannon Research Institute, Nashville, TN

Corresponding Author: Cheng Seok Quah (cquah@arcusbio.com) ASCO GI Virtual Congress 2021

Abstract #404

WT MUT

Dose-Escalation Dose-Expansion

100 mg AB680 IV Q2W +NP/Gem + Zim

AB680 RDE +NP/Gem + Zim

AB680 RDE +NP/Gem + Zim

AB680 RDE +NP/Gem

75 mg AB680 IV Q2W +NP/Gem + Zim

50 mg AB680 IV Q2W +NP/Gem + Zim

25 mg AB680 IV Q2W +NP/Gem + Zim

1L metastaticPDAC

1L metastaticPDAC

Safety monitoring throughout treatment period; radiographic disease evaluation every 8 weeks

R2:1

Pending futility analysis

Identification of AB680 RDE +NP/Gem + Zim